Development and validation of warning system of ventricular tachyarrhythmia in patients with heart failure with heart rate variability data

788 patients with mild to moderate chronic, stable heart failure (NYHA class II or III, LVEF ≤35%) enrolled in the ICD arm of the Sudden Cardiac Death–Heart Failure Trial (SCD-HeFT). Median age 60, 609 males.

Machine learning algorithms (Random Forest and Support Vector Machine) using heart rate variability (HRV) features extracted from ICD data to predict ventricular tachyarrhythmia 5 minutes and 10 seconds before onset.

Regular heart rhythms compared to pre-ventricular-tachyarrhythmic rhythms (pre-appropriate-shock rhythms) from the same patient cohort.

Area under the receiver operating characteristic curve (AUC) for the prediction of ventricular tachyarrhythmia at 5 minutes and 10 seconds prior to onset.